causallib.estimation.tmle module
- class causallib.estimation.tmle.CleverCovariateFeatureMatrix(weight_model)[source]
Bases:
causallib.estimation.tmle.CleverCovariateImportanceSamplingMatrix
Clever covariate uses a matrix of inverse propensity weights of all treatment values as a predictor to the targeting regression.
References
Gruber S, van der Laan M. tmle: An R package for targeted maximum likelihood estimation. 2012. https://doi.org/10.18637/jss.v051.i13
- class causallib.estimation.tmle.CleverCovariateFeatureVector(weight_model)[source]
Bases:
causallib.estimation.tmle.BaseCleverCovariate
Clever covariate uses a signed vector of inverse propensity weights, with control group have their weights negated. The vector is then used as a predictor to the targeting regression.
References
Schuler MS, Rose S. Targeted maximum likelihood estimation for causal inference in observational studies. 2017 https://doi.org/10.1093/aje/kww165
- class causallib.estimation.tmle.CleverCovariateImportanceSamplingMatrix(weight_model)[source]
Bases:
causallib.estimation.tmle.BaseCleverCovariate
Clever covariate of inverse propensity weight vector is used as weight for the targeting regression. The predictors are a one-hot (full dummy) encoding of the treatment assignment.
References
Gruber S, van der Laan M. tmle: An R package for targeted maximum likelihood estimation. 2012. https://doi.org/10.18637/jss.v051.i13
- class causallib.estimation.tmle.CleverCovariateImportanceSamplingVector(weight_model)[source]
Bases:
causallib.estimation.tmle.BaseCleverCovariate
Clever covariate of inverse propensity weight vector is used as weight for the targeting regression. The predictors are a signed vector with negative 1 for the control group.
- class causallib.estimation.tmle.TMLE(outcome_model, weight_model, outcome_covariates=None, weight_covariates=None, reduced=False, importance_sampling=False, glm_fit_kwargs=None)[source]
Bases:
causallib.estimation.doubly_robust.BaseDoublyRobust
Targeted Maximum Likelihood Estimation. A model that takes an outcome model that was optimized to predict E[Y|X,A], and “retargets” (“updates”) it to estimate E[Y^A|X] using a “clever covariate” constructed from the inverse propensity weights.
- Steps:
Fit an outcome model Y=Q(X,A).
Fit a weight model A=g(X,A).
Construct a clever covariate using g(X,A).
Fit a logistic regression model Q* to predict Y using g(X,A) as features and Q(X,A) as offset.
Predict counterfactual outcome for treatment value a Q*(X,a) by plugging in Q(X,a) as offset, g(X,a) as covariate.
Implements 4 flavours of TMLE controlled by the reduced and importance_sampling parameters. importance_sampling=True moves the clever covariate from being a feature to being a sample weight in the targeted regression. ‘reduced=True’ use a clever covariate vector of 1s and -1s, therefore only good for binary treatment. Otherwise, the clever covariate are the entire IPW matrix and can be used for multiple treatments.
References
TMLE: Van Der Laan MJ, Rubin D. Targeted maximum likelihood learning. 2006. https://doi.org/10.2202/1557-4679.1043
TMLE with a vector version of clever covariate: Schuler MS, Rose S. Targeted maximum likelihood estimation for causal inference in observational studies. 2017. https://doi.org/10.1093/aje/kww165
TMLE with a matrix version of clever covariate: Gruber S, van der Laan M. tmle: An R package for targeted maximum likelihood estimation. 2012. https://doi.org/10.18637/jss.v051.i13
TMLE with weighted regression and matrix of clever covariate: Gruber S, van der Laan M, Kennedy C. tmle: Targeted Maximum Likelihood Estimation. Cran documentation. https://cran.r-project.org/web/packages/tmle/index.html
TMLE for continuous outcomes Gruber S, van der Laan MJ. A targeted maximum likelihood estimator of a causal effect on a bounded continuous outcome. 2010. https://doi.org/10.2202/1557-4679.1260
- Parameters
outcome_model (IndividualOutcomeEstimator) – An initial prediction of the outcome
weight_model (PropensityEstimator) – An IPW model predicting the treatment.
outcome_covariates (array) – Covariates to use for outcome model. If None - all covariates passed will be used. Either list of column names or boolean mask.
weight_covariates (array) – Covariates to use for weight model. If None - all covariates passed will be used. Either list of column names or boolean mask.
reduced (bool) – If True uses a vector version of the clever covariate (rather than a matrix of all treatment values). If True enforces a binary treatment assignment.
importance_sampling (bool) – If True moves the clever covariate from being a feature to being a weight in the regression.
glm_fit_kwargs (dict) – Additional kwargs for statsmodels’ GLM.fit(). Can be used for example for refining the optimizers. see: https://www.statsmodels.org/stable/generated/statsmodels.genmod.generalized_linear_model.GLM.fit.html
- estimate_individual_outcome(X, a, treatment_values=None, predict_proba=None)[source]
Estimates individual outcome under different treatment values (interventions)
- Parameters
X (pd.DataFrame) – Covariate matrix of size (num_subjects, num_features).
a (pd.Series) – Treatment assignment of size (num_subjects,).
treatment_values (Any) – Desired treatment value/s to use when estimating the counterfactual outcome/ If not supplied, calculates for all available treatment values.
predict_proba (bool | None) – In case the outcome task is classification and in case learner supports the operation, if True - prediction will utilize learner’s predict_proba or decision_function which returns a continuous matrix of size (n_samples, n_classes). If False - predict will be used and return value will be based on a vector of class classifications. If None - parameter is ignored and behaviour is as specified when initializing the IndividualOutcomeEstimator.
- Returns
- DataFrame which columns are treatment values and rows are individuals: each column is a vector
size (num_samples,) that contains the estimated outcome for each individual under the treatment value in the corresponding key.
- Return type
pd.DataFrame
- fit(X, a, y, refit_weight_model=True, **kwargs)[source]
Trains a causal model from observed data.
- Parameters
X (pd.DataFrame) – Covariate matrix of size (num_subjects, num_features).
a (pd.Series) – Treatment assignment of size (num_subjects,).
y (pd.Series) – Observed outcome of size (num_subjects,).
sample_weight – To be passed to the underlining scikit-learn’s fit method.
- Returns
A causal weight model with an inner learner fitted.
- Return type
- class causallib.estimation.tmle.TargetMinMaxScaler(feature_range=(0, 1), *, copy=True, clip=False)[source]
Bases:
sklearn.preprocessing._data.MinMaxScaler
A MinMaxScaler that operates on a vector (Series)
- fit(X, y=None)[source]
Compute the minimum and maximum to be used for later scaling.
- Parameters
X (array-like of shape (n_samples, n_features)) – The data used to compute the per-feature minimum and maximum used for later scaling along the features axis.
y (None) – Ignored.
- Returns
self – Fitted scaler.
- Return type